The clinical utility for autonomic research purposes of the Finapres, a device for measuring finger arterial pressure continuously and noninvasively, was assessed by estimating its accuracy, precision and within-subject variability in 48 subjects, aged 18-65 years, in comparison with intraarterial brachial blood pressure. At differences of -3.5 +/- 12, -8.0 +/- 8 and -4.4 +/- 8 mmHg from simultaneous intrabrachial systolic, mean and diastolic pressure, the Finapres meets the Association for the Advancement of Medical Instruments (AAMI) accuracy requirements for systolic and diastolic, but not for mean pressure. The precision requirements were nearly met for mean and diastolic, but not for systolic pressure. These results compare to those of others under widely varying circumstances in anaesthetized and awake subjects, and are comparable to those published for the auscultatory technique. The within-subject precision of Finapres is high; the 95% confidence intervals are 3.4-4.5 mmHg for systolic, 1.5-2.0 mmHg for mean and 1.7-2.2 mmHg for diastolic. This makes the Finapres suitable for tracking changes in blood pressure. Four case studies are provided as examples of the value of the Finapres in the clinical laboratory.